Starslayer and I were talking about this the other day. I have to admit that I like what they did in GSF (4thE or whatever it is called) with the tech tree concept. The details are a nightmare, and largely I think most of the Threat-Reaction-Percieved-Threat-to-Threat-From-Threat rules should have been put in as optional only to be used with a SM since most of them are there to prevent the player doing some rules lawyering to develop stuff faster then normal. But otherwise the knot system does a good job at keeping the technology a lot less cookie cutter then it is in 3rdR. Unfortunately the weapons became pretty much generic from the push to balance all the tech trees in my view.
Our discussion moved on to our games tech development. Now we have some drastically different house rules in play. Shipyard rates are halved: so a SY builds at 5+TL, and our growth is basically the per turn growth but applied ever 10 months (so we undid the artifical 10x difference between economic and military activities to a large degree). The result seems to be a game much closer to the official history from Stars At War then is typical of most starfire campaigns. But we still kept the tech rules the same. The result is generally speaking you finish one prototype and your tech level goes up and you start the next prototype and so you are largely a generation behind ship wise. The SCN has all of 2 BBs but I just hit TL7 and can start a prototype DN if I wanted. Prototypes at 2.5+TL/2 build rates take a long time.
Another aspect is that SM2 largely didn't change anything from Imperial Starfire so we are still using 3rdE tech development rules with a few minor changes. The rate at which you can do EL research is determined by the rate at which you can develop 25% of the tech systems for a tech level. While that time is somewhat random the reality is that it approximately the same at TL1 as TL20, I would say it is likely faster at TL20 then it is at TL1 due to the increased ability to afford Tech research.
Developing a tech system cost money and time. The money is only relevant at the lower tech levels due to the economic nightmare that is Starfires compound interest economy. As your EL goes up and time passes what was once expensive becomes trivial. SM2 corrects for this by assigning a minimum 10% EL development cost (and minimum 30% start up fee) but except for an NPR no player race will blink at that. NPRs could potentially get into a situation where their maintenance cripples development of new technologies. But for a player race the first 4 Techlevels are painfuly as the real cost of EL research is a substantial fraction of your economic output, but then you get a substantial break from TL4-TL8 where the cost realtive to your economy drops (at this point I pay 10% for example and I hardly notice it). So basically whatever logic went into the table in Imperial Starfire and I have to assume that there was some reason for it...well it is clear in the SM2 economic reality it no longer fits well. I believe that income growth under SM2 is substantially higher than under ISFs rules, given how trivial colonization has become. So there is no monetary break on EL research, it contributes eventually to the "Rich get Richer, Faster and Faster" syndrome.
The cost in research points is also set by some arcane formula that isn't at all clear why it is like it is. In reality based on some stuff C.J. Cherryh once put up on her website actual technological development is a complex theme and so I don't know how to do anything with those numbers. I have to admit some sort of base+(EL_underdevelopmentx(factor)) would appeal to me more. Uplifting should be handled very differently since it non-sensically requires the race to progress through all the interveening techlevels which isn't an "uplift" to my thinking.
The real point that determines when you get your new EL (outside of random die luck) is how long it took you to develop 25% of the last ELs tech systems. That is a factor of how many were there...as the 25% requirement basicaly means devop 2-5 systems and start a new EL. So in principle it comes down to random luck and random luck and arbitrary numbers.
How to fix this is unclear or even if it matters enough to need fixing.
Then you get to tech system development. At this point I have to admit I find the costs of the various tech systems assigned at what seems at random. There is some effort to make game changing systems expensive but this is no applied uniformly, plus due to the cludging in of disseperate systems you get a potpuri of wierd. But ultimately if any price of any tech system makes sense is crushed by the weight of the starfire compount interest economics. In the end it doesn't matter, the player can afford it. The prices if viewed from the perspective of an EL1 race may be extreme but when you are there they are not so. The RM for example started development of all fighter systems F0, V, fG, and fR in 2 turns. When they hit EL9 they will pay for in the first turn F1, fL and fM, after that I need to think. But this is only an issue due to them having a fleet maintence of 70% of their economy (and that will drop when they conquor the Buer, as it has done a serious nose dive due to their two trade treaties). And our incomes are comparitively low, for most campaigns by the time they reach our stage money isn't an issue and hasn't been for ages. So the cost of the system largely isn't an issue in the game. Steve's Rigilians were throwing hundreds of thousands of MCr at advanced R&D without blinking. Kurt may or many not have the same issue I believe the Terran Empire's cost in tech expenditure outside of EL research is less then the governments toilet paper budget.
Speaking for myself I would do something to a lot of the costs since a fair few of the are just plain wierd but this is a juggling issue. I can't fathom what was the original basis for the costs of things so to re-do the costs you basically have to start from nothing and develop your own pricing system. Which as I point out above me eventually becomes the next best thing to meaningless. What I would change is the time to develop it. I would go to a system of cost in research points is 23+(2xTL_of_system). So it take 25 pts at TL1 and 43 at TL10. I would then, admittedly arbitrarily rate a system as to how much developing it changes the game. I would then apply a multiplier to the basic research costs. What that multiplier is depends on what the effect of the system is, or in other words it would take play testing. The "2" in the above forumla is just a initial guess, it probably isn't the right number but as a start point it is ok.
One other thing when you add in all the TRPTPTTP stuff what it doesn't do it increase the basic 25 pt development time and when combined with money is not relevant at some stage in the game you get the player able to develop technology that can be game shattering at a cost they barely blink to pay. I can get ? for 25,000 MCr at TL5??? wow..where do I sign the check? Ultimately Starfire's reliance on money to limit things just fails, it can't work in the long run because the economic system is nothing more than compound interest accounting. This is largely due to the removal in SM2 of virtually all maintence costs for anything other than your fleet plus the inclusion of the CFN rules. This makes the cost to run an empire largely independant of the size of the empire and that means "The Rich get Richer, Faster and Faster." Crew points were there to keep fleet sizes reasonable...remove them and fleet size explodes. The rules for emplacement time and the fact that it meant a reduction in the sending planets GPV were there to keep colonization in check remove them add in the CFN and colonization explodes. There may have been some point in terms of checks and balances with the tech systems via transmitting data and TL's of shipyards and such but that also got axed.
The goal of all these changes was to simplify book keeping but the balooning of fleet sizes and masive colonization that occurs (to the point of every rock in every known system being settled) increases book keeping as well. Not to mention tracking the tens of thousands of patterns, buoys etc that their lack of maintence encourages. Given I have written spreadsheets to handle both 3rdR and GSF games I am continously impressed by Steve's work on SFA. But the simple fact is that even under GSF my spread sheet for known systems was already at turn 20 or so 200 lines long (each system takes 10-15 lines), scrolling through it and updating that was a pain. I had 3 pages for ships and fleets and shipyards and my ship list was also a few hundred lines long. Without a proper database program (EXCEL is not that) I am actually of the opinion that doing it by hand with a binder and calculator would be not much more trouble.
Anyway the last two points are not so relevant to what I want to say. In summary I'd suggest anyone looking to do a 3rdR campaign consider adopting a more GSF EL research idea. Where you are basically continuously developing EL at 10% of your income (rather than the way it is done now where the penalizing cost is for the first few EL jumps). I would also suggest chaning the tech system development cost in points to something that makes higher tech systems cost more of them.